The benzodiazepines (BZs; e.g. Valium) are widely used clinically for their anti-anxiety, sedative-hypnotic, anti-convulsant, and muscle-relavant effects, although the mechanisms underlying these effects are poorly understood. Tolerance and physical dependence occur with prolonged use; again, the mechanisms are unclear. The discovery of specific CNS BZ receptors stimulated an intense interest in this area since a clear understanding of BZ mechanisms would facilitate the design of analogs with greater specificity of actions. Genetic studeis can provide useful information about drug action. In a pilot study mass selection was used to breed divergent lines of mice which are resistant and sensitive to diazepam (DZ). Mice were selected for mating based on the duration of impairment on the rotarod (i.e. ataxia) after a standard diazepam dose of 20 mg/kg; by the seventh generation there was no overlap in the two lines. These mice have been examined for: a) CNS senistivity to DZ; b) DZ absorption and elimination; c) sensitivity to other BZs, GABA agonists, and sedative hypnotics; d) BZ receptor density and affintiy in various brain areas; e) ethanol tolerance and physical dependence; and f) chloride influx at the BZ-GABA receptor complex. Mass selection leads to rapid deivergence, but inbreeding may limit the value of these animals for subsequent long-term studies. I therefore propose to repeat the previous work using state-of-the -art genetic procedures, i.e., within-family selection with replicate bi-directional lines and contemporaneous control lines. This method minimizes inbreeding and leads to the development of lines which ideally differ only in genes related to the observed BZ response. The resulting divergent lines will constitute a valuable experimental resource for the study of BZ mechanisms. First, we will obtain the first information regarding the genetics of BZ action. Second, we will investigate neurologic differences between the two lines which might identify the mechanisms of action. Third, continued studies with ethanol and barbiturates will provide valuable information about mechanisms of cross- tolerance and -dependence. And finally, studies on the co- segregation of other BZ actions in the two lines will provide evidence about the commonality of underlying mechanisms.